As I noted in my P.P.S. above, I agree that seats are the relevant unit for evaluating political strategy and for considering policy implications of the election. But I think vote share is the relevant measure when studying public opinion, in particular in asking whether voters were trying to balance their Obama support with Republican voting for the House.

I agree that the roll-off in downballot races is important; some comes in elections that are uncontested or not seriously contested. Beyond this, we know from polls that voters don’t generally have positive feelings about Congress in any case.

This analysis is very interesting, but it misses the point. As the saying goes “the proof of the pudding is in the eating” and the Democratics did worse than expected in picking up seats in Congress. This is all the more surprising, because the presidential polls were quite accurate, both overall and state-by-state.

The analysis should be done by Congressional district, rather than by state. The logic is simple: the electoral votes for president (for the most part) are winner-take-all at the state level, so states are the important unit for analysis. Congressional seats are winner-take all at the district level, so districts are the corresponding important unit for analysis.

If the additional votes for Democrats came in districts that are overwhelmingly Democratic, then the national vote totals change, but the impact on Congress is much less.

Very interesting to me are the facts that there were 6.8M fewer votes for Democrats in House races than for Obama, and 8.4M fewer votes for Republicans than for McCain. Many people voted for the presidential candidates but did not support that party for Congress.

These numbers are actually larger than in 2004, when the difference was about 6.3M for both parties. This suggests that Obama’s “coattails” were no stronger than Kerry’s (this is only a suggestion because we do not know the actual proportion of party splitting). However, McCain’s were considerably worse than Bush or Obama.

–gordon

I think you’ve made your overall point nicely, but your charts also reveal that over a 50 year span the public has favored a center-left Democratic congress and an authoritarian-leaning executive. More evidence that Obama’s achievement is pretty notable.

This GOP strategy of gerrymandering districts to give them narrow majorities in most districts only works if the overall vote is close. In 2006, it wasn’t, so they lost a large number of seats.

Perhaps the reason that the GOP House losses of seats were considered not so bad compared to 2006 was because in 2008 the Democrats ran up huge turnouts in black-represented Congressional districts, which were already all Democratic? That’s something you might look into.